Video & Animation

Use when someone shares a reference video, ad, Reel, Short or TikTok (link or file) and wants their own video to follow its structure, pacing or visual style

Fetch film and TV series data from Filmweb.pl: search by title, rating and number of votes, critics rating, genres, duration, directors, cast, plot, premiere dates, where to watch (VOD providers and prices) and actor details. Use whenever the user asks about a movie or series on Filmweb, pastes a filmweb.pl link, or wants a Polish rating, opis, obsada or VOD info, e.g. "ile Matrix ma na Filmwebie", "kto grał w…", "kto wyreżyserował…", "o czym jest…", "gdzie obejrzę…", "kiedy premiera…", "jaki serial…", "ocena serialu…".

Optimize 3D assets (.glb and .gltf) and integrate them in interactive web animation with Three.js, React Three Fiber, GSAP, and gltf-transform. Use for compression (Draco, Meshopt, KTX2), pivot and animation setup, and web rendering.

Create a new visual style for the creator's videos in a Reel Studio workspace (styles/<id>/style.json): interview him for its name, feel, colours and what each one means, fonts, motion, where pictures come from, reference links and images, then write the style so it shows in Estilos and new videos can use it. Use when he asks for a new style or presses "Nuevo estilo" on the Estilos page.

当用户在做 AI 编程/多分支并发开发、出海工具站 SEO/GEO 增长、AIGC 视频与代码渲染动效（Video as Code）、消除代码/UI/视频/写作中的「AI 味」、大模型与 Agent 性能实测选型、或传统企业 FDE 交付与一人公司自动化运营时调用。浓缩 20 位中文 X 技术创作者 1,145 条实战帖的六大可执行工作流：Agent 软件工厂、SEO/GEO 增长引擎、Video as Code 视觉流水线、全栈去 AI 味协议、产物级 Benchmark 与模型路由、FDE 交付与高管审核制。不适用于：实时股价/币价查询、与 AI/出海/内容生产无关的纯行政流程、要求绕过支付校验或人工闸门的高危操作。Triggers: AI 创作者方法论/软件工厂/worktree 并发/xcode 冷编译/一词一页/TDH/GEO 意图组/Fan-out 选题/白模预演/Video as Code/去 AI 味/早期智商债/FDE 交付/高管审核制。

The "Persona hablando + animaciones" format of a Reel Studio video — the creator talks to camera in his own recording; you transcribe it, agree the cuts with him (pauses, flubs, retakes), cut it, and lay the video's style animations over or between his shots. Use for any video whose format.txt says persona-hablando.

Write, polish, or audit short Korean social posts (Threads, X, Instagram captions, community posts) so they read like a real person telling a true story, not like AI. Use when the user asks for a Threads post, 스레드 글, 바이럴 글, 후킹 글, 일상 글, SNS 글, 썰, wants hooks for a Korean post, says a draft "AI 같다", "너무 무겁다", "뭔 소린지 모르겠다", or wants a draft rewritten in natural 구어체. Combines hook craft, storytelling, Korean anti-AI editing, and spoken-register rules into one pass, with real material required before drafting.

Cria vídeo de motion no estilo explicador de produto SaaS, com interface real em vidro, luz líquida verde reagindo à música, câmera 3D rápida, revelação da logo com corte de lâmina, narração, trilha e SFX sincronizados e pacote para o DaVinci. Na primeira vez, conhece a pessoa e ensina a usar em 8 passos. Também melhora um vídeo que ela enviar e aceita outros vídeos como referência. Use sempre que a pessoa pedir para melhorar um vídeo, adicionar referências, um motion ou vídeo "nesse estilo", "igual ao Wakizashi", "com vidro", um vídeo novo sobre o tema da pessoa, ou mudanças de um vídeo feito assim, mesmo sem citar a skill. Não use para editar vídeo gravado com apresentador; nesse caso, ofereça a skill Fluxo de edição, descrita em references/dependencias.md.

Profile a YouTube channel from the official YouTube Data API v3. Covers subscribers, views, upload cadence, recent views and trend, engagement, Shorts share, top videos and an estimated ad earnings range with its formula. Use when given a YouTube channel URL, @handle or channel ID and asked how big, active, engaged or valuable the channel is.

Make a series of narrated-by-callout training videos for any web system, recorded from the running system with a slow cursor, click rings, arrow callouts, smooth zooms, a step counter, sound effects (clicks, typing, whooshes), chapter videos and one full video with chapter markers. The colours, fonts, corners and logo are read from the system itself, so every system gets its own look in the same proven layout. Use this whenever someone wants training videos, a video tour, a walkthrough, screen-recorded tutorials, how-to or onboarding videos for an app, portal, dashboard or internal system, even if they only say "make videos for this system", "record how it works", or "فيديوهات تدريب للنظام". Not for editing existing footage, marketing promos, or desktop apps.

Use when cutting or trimming video, adding titles and graphics, grading colour, mixing audio, exporting MP4/GIF or stills, when running filmcraft-cli, or when connecting to the FilmCraft app (storytold) on this machine

Write short-form video scripts that hold viewers, using the creator's own video analytics. Use when choosing the hook of a chosen idea, writing or tightening a script before filming (talking-head or narrated explainer), or turning a posted video's insights into lessons. Coming up with ideas is content-ideas'.

Summarize a video through the VSS Pipeline Manager - start a summary pipeline with POST /summary (full required body), poll GET /summary/{stateId} until complete, then return the summary via GET /summary/{stateId}/raw. Use when the user says "summarize this video", "create a summary", "what happens in this video" (on an ingested video), or wants to run/inspect the summarization pipeline. Requires a summary-capable deployment (--summary, --dual, or --unified).

Trim the fat from raw talking footage. Turns raw video into a tight first cut by removing dead air, ums and uhs, stutters, false starts and bad takes (keeping the last good take), then checks the result by listening to it again. Use when the user says "rough cut", "first cut", "trim the fat", "cut the pauses", "remove the ums", "cut my retakes", "make it snappy", or hands over raw footage to edit.

Build an audited "co-reading" explainer — an interactive isometric figure (Anatomy kit) wrapped as an anywidget for Jupyter / marimo, with 1920×1080 stills, a short two-voice video script, and a ledger of what can / cannot be said — from the frozen outputs of a pre-registered measurement (a "card" = PREREG + script + JSON with provenance), so that a human reader and an AI read the same readout and align their understanding with what the instrument can read. Use this skill whenever the user asks for 図解 / 立体図 / 説明図 / 「anatomy*anywidget」/ an explainer or figure of an instrument readout (logit or tuned lens, probe, benchmark run, sensor log), a figure "for the editor / reviewers", material for ゆっくり動画, a side-by-side of a success and a failure case, or wants to turn a question raised by a figure into a frozen check ("この図で…を確かめたい", "どこから…が湧くのか") — even without saying "co-reading". Layered on the `anatomy` skill (MIT). Not for a quick one-off chart (use dataviz) or a decorative illustration with no data behind it.

How a narrated explainer video is made in a Reel Studio workspace, in any visual style — script, pictures, voice, images, build, board, sound, mute test, render — following the creator's settings from the studio (profile, format, voice, image sources). Use for any new or revised video in a project's videos/ (projects/<id>/videos/). The video's style adds its own skill (styles/<id>/SKILL.md) for the look and the Remotion kit; LESSONS.md has what past videos taught.

Automated audio extraction, Whisper speech-to-text, Luna dual-verification lyrics alignment, local Edge LLM proofreading, CSS subtitle styling, and NVENC GPU video hardsub burn-in. Activate this skill when processing video/audio for subtitles, aligning lyrics, running batch subtitle automation, or configuring ASS/CSS hardcopy video burning.

Agente Produtor de Vídeos em Código via Remotion & Multi-Formato. Especialista em criação programática de vídeos com React, animações sincronizadas com áudio, legendas automáticas, teasers de SaaS, slides dinâmicos e renderização automatizada.

Make animated videos in code with Remotion (React, SVG, three.js): explainers, character stories, motion graphics, kinetic typography, paper collage and animated infographics, in 2D or 3D, from a script or voiceover, in any aspect ratio. Use when asked to animate a script, make an animated video or Reel, or motion graphics. Not for editing camera footage (video-edit) or AI-generated clips (higgsfield-*).

Builds figures that move or respond to the reader, in three tiers: SMIL animations and slider widgets inside lessons, interactive mechanism figures that show the named parts of a real system with a computed verdict, and ordered dither bands in the course palette. Triggers: "animated figure", "add a slider widget", "interactive figure", "show how this system works", "make this diagram move", "dither masthead", "mechanism figure", "figure fence".

Coach the user through a learning-project task from start to finish - brief the task with an interactive visual explainer (diagrams, animation, simulator), plan small checkpoints, guide each step with a hint ladder instead of ready answers, coach debugging, verify against the task's "Verified result", run the control questions one at a time, and close with a summary. Use this whenever the user works on a course or self-study task list (T01, T07, "next task", "lets start next task", "start T08"), asks to explain a concept from the material, says they don't know an answer, answers a control question, or wraps up a task - even if they don't mention learning or coaching explicitly. Also use it when the user asks "do it for me" inside such a task, so the code still comes with an explanation.

Design, refine or critique web UI, landing pages, decks and image sets so they read as chosen, not generated. Use for a new page, a redesign, a design review, polish, motion or a brief.

Complete guidelines for designing effective, beautiful presentations: message-first process (audience, goal, script, storyboard, slide count), deck structure and slide types (cover, index, break, focus, content, back cover with CTA), layout (margins, grids, alignment, proximity, hierarchy, focus by dimming), a type scale for 1920×1080 slides, message titles, bullets and their alternatives, color palettes and roles, images, icons, backgrounds, shadows and text-on-photo, charts, tables, big numbers and timelines, animation and morph principles with timings, contexts (school, thesis, corporate, pitch deck, product launch, portfolio, training, talks, webinars), a before→after redesign method, templates, brand and AI. Use this skill whenever you create, generate, review, redesign or critique a presentation, slide deck, pitch deck, PowerPoint, Keynote, Google Slides or Canva deck (including generating .pptx with code or DeckCraft), or advise on slides, even if the user only says 'make me a ppt'.

Author, animate, revise, inspect, score, reformat, render or deliver editable native Carvercat motion graphics through its MCP tools - kinetic type, transitions, lower thirds, logo reveals, charts, UI motion, sound design and 9:16 versions - with a verified pattern library and craft rules. Use for app operation and creative graphics, not Carvercat source implementation.